Summary of Using Degeneracy in the Loss Landscape For Mechanistic Interpretability, by Lucius Bushnaq et al.
Using Degeneracy in the Loss Landscape for Mechanistic Interpretabilityby Lucius Bushnaq, Jake Mendel, Stefan Heimersheim,…
Using Degeneracy in the Loss Landscape for Mechanistic Interpretabilityby Lucius Bushnaq, Jake Mendel, Stefan Heimersheim,…
Function Extrapolation with Neural Networks and Its Application for Manifoldsby Guy Hay, Nir SharonFirst submitted…
Monitizer: Automating Design and Evaluation of Neural Network Monitorsby Muqsit Azeem, Marta Grobelna, Sudeep Kanav,…
Parallel Backpropagation for Shared-Feature Visualizationby Alexander Lappe, Anna Bognár, Ghazaleh Ghamkhari Nejad, Albert Mukovskiy, Lucas…
A Machine Learning Approach for Simultaneous Demapping of QAM and APSK Constellationsby Arwin Gansekoele, Alexios…
Modeling Bilingual Sentence Processing: Evaluating RNN and Transformer Architectures for Cross-Language Structural Primingby Demi Zhang,…
Spectral complexity of deep neural networksby Simmaco Di Lillo, Domenico Marinucci, Michele Salvi, Stefano VigognaFirst…
Error-margin Analysis for Hidden Neuron Activation Labelsby Abhilekha Dalal, Rushrukh Rayan, Pascal HitzlerFirst submitted to…
Agnostic Active Learning of Single Index Models with Linear Sample Complexityby Aarshvi Gajjar, Wai Ming…
Neural Active Learning Meets the Partial Monitoring Frameworkby Maxime Heuillet, Ola Ahmad, Audrey DurandFirst submitted…